Elizabeth A. Saffer

Five-year results of a randomized clinical trial comparing total mastectomy and segmental mastectomy with or without radiation in the treatment of breast cancer

In 1976 we began a randomized trial to evaluate breast conservation by a segmental mastectomy in the treatment of Stage I and II breast tumors less than or equal to 4 cm in size. The operation removes only sufficient tissue to ensure that margins of resected specimens are free of tumor. Women were randomly assigned to total mastectomy, segmental mastectomy alone, or segmental mastectomy followed by breast irradiation. All patients had axillary dissections, and patients with positive nodes received chemotherapy. Life-table estimates based on data from 1843 women indicated that treatment by segmental mastectomy, with or without breast irradiation, resulted in disease-free, distant-disease-free, and overall survival at five years that was no worse than that after total breast removal. In fact, disease-free survival after segmental mastectomy plus radiation was better than disease-free survival after total mastectomy (P = 0.04), and overall survival after segmental mastectomy, with or without radiation, was better than overall survival after total mastectomy (P = 0.07, and 0.06, respectively). A total of 92.3 per cent of women treated with radiation remained free of breast tumor at five years, as compared with 72.1 per cent of those receiving no radiation (P less than 0.001). Among patients with positive nodes 97.9 per cent of women treated with radiation and 63.8 per cent of those receiving no radiation remained tumor-free (P less than 0.001), although both groups received chemotherapy. We conclude that segmental mastectomy, followed by breast irradiation in all patients and adjuvant chemotherapy in women with positive nodes, is appropriate therapy for Stage I and II breast tumors less than or equal to 4 cm, provided that margins of resected specimens are free of tumor.

Studies concerning the regional lymph node in cancer. IV. Tumor inhibition by regional lymph node cells

Observations in two different syngeneic tumor‐host systems utilizing in vivo neutralization and in vitro cytotoxicity have revealed that even after tumors had been present for a prolonged period of time, regional lymph node cells were capable of interfering with the growth of tumor cells. Neither distant LNCs nor spleen cells ever fully displayed that characteristic. Only when animals approached death from their tumors did RLNCs demonstrate loss of their neutralizing capability. Inhibition of growth of tumor cells by RLNCs was not impaired for at least as long as 2 months following removal of primary tumors. Cells obtained at that time from other sources failed to display such a capability. The findings are compatible with others from this laboratory which have indicated that RLNs are unique from the rest of the lymphoreticular system insofar as the immunologic response of a host to its tumor is concerned. They also suggest that disseminated living tumor cells gaining access to RLNs may be destroyed in those structures, and that the finding of negative lymph nodes may be the result of such a circumstance rather than that a tumor had been removed prior to its lymphatic dissemination.

Inhibitory effect of prolonged corynebacterium parvum and cyclophosphamide administration on the growth of established tumors

The present study was carried out to investigate the effects of prolonged administration of C. parvum alone and in combination with cyclophosphamide for the treatment of established, measurable C3H tumors. The continued weekly administration of C. parvum by itself provided a limited but significant inhibitory effect on tumor growth and significantly prolonged survival. Intraperitoneal and intravenous administration was found to be more effective than the subcutaneous route. When C. parvum was administered asynchronously in combination with cyclophosphamide at weekly intervals a tumor growth inhibitory effect was achieved which was greater than that resulting from either agent alone. Such an effect was consistently obtained and was seemingly independent of the sequence of drug administration. When cyclophosphamide preceded the initial C. parvum administration, arrest in the rate of tumor growth occurred, resulting in infinite tumor doubling time for the duration of observation (>90 days). The combination of C. parvum and cyclophosphamide produced a more effective inhibition of tumor growth than did BCG and cyclophosphamide similarly employed. The importance of these findings relative to clinical application is considered. While the significance and genesis of the marked desmoplastic reaction characterizing tumors from animals treated with C. parvum and cyclophosphamide is at present speculative, consideration is given to the possibility that this could signify a host response against tumor growth.

Studies concerning the regional lymph node in cancer. III. Response of regional lymph node cells from breast and colon cancer patients to PHA stimulation

To obtain information regarding the immunologic capacity of regional lymph nodes (RLNs) in patients with cancer, investigations were performed utilizing lymph nodes and/or blood lymphocytes from 140 patients with operable breast or colon cancers, benign breast disease, or cardiac problems requiring open heart surgery. Cells of all regional lymph nodes (RLNCs) from breast cancer patients responded to PHA stimulation. The response of stimulated and non‐stimulated RLNCs from patients with negative nodes was significantly greater than that of cells from women with positive nodes. The presence or absence of tumor in nodes from positive node patients had no effect on 3HT uptake by RLNCs from such nodes. In patients with breast cancer, regardless of nodal status, a significant variation in 3HT uptake was observed between cells derived from different RLNs within the same patient. A similar variability occurred between patients. Such findings were noted when RLNCs were cultured with or without PHA. There was no significant difference in uptake of 3HT by peripheral blood lymphocytes (PBLs) from breast cancer patients and women with benign breast disease. When PBLs and RLNCs from the same breast cancer patients were exposed to undiluted PHA, the response of the former was significantly greater than that of the latter. While both types of cells were less stimulated by diluted PHA, the response by RLNCs was significantly greater. Results from colon cancer patients were similar to those from patients with breast cancer. Virtually all nodes responded to PHA stimulation and the 3HT uptake by RLNCs was greater when they came from negative rather than positive node patients. There was a variation in response of nodes to PHA in individual patients and between patients. Unlike findings with breast cancer, the response of RLNCs and PBLs from patients with colon cancer to undiluted PHA was not significantly different, but as with the former disease, RLNCs demonstrated a greater response to dilute PHA. RLNCs from patients with breast and colon cancer were stimulated by PHA to a greater degree than LNCs from patients undergoing open heart surgery. Whether this difference indicates that cells from patients with heart disease are less responsive, or that those from cancer patients are more so is uncertain. In conclusion, the present studies indicate that virtually all RLNs (with and without metastases) from patients with primary operable breast and colon cancers contain cells capable of responding to PHA stimulation. If, as is generally considered, such a response is indicative of lymphocyte immunocompetence, then RLNs continue to possess immunologic capabilities despite the presence of growing tumors.

Studies concerning the regional lymph node in cancer. VI. Correlation of lymphocyte transformation of regional node cells and some histopathologic discriminants

Regional nodes from patients with mammary carcinoma exhibited greater sinus histiocytosis and less lymph follicles than those from persons with colonic carcinoma. These latter tumors contained more lymphoid infiltrate, and their component cells had higher nuclear grades than breast carcinomas. Differences in the relationship of these histopathologic parameters with lymphocyte transformation of RLNCs in these two tumor types indicates the singularity of neoplasms of diverse origins. Lower nuclear grade and mild lymphoid infiltrate of mammary carcinoma were associated with nodal lymph follicle formation and the converse with the absence of such nodal structures. No significant differences in these histopathological discriminants were observed in patients with or without nodal metastases. Lymphocyte transformation of RLNCs from patients with mammary carcinoma was increased in those in whom their carcinomas were of low nuclear grade and contained only mild degrees of lymphoid infiltrate. These findings suggest that nuclear grade and lymphoid infiltrates may be differently related to tumor host responses than previously contended. No functional or other relationships could be discerned for sinus histiocytosis. Ultrastructurally, cells comprising this latter lesion appeared as banal macrophages. Activated lymphocytes appeared to represent the morphological analog of lymphocyte transformation.

Studies concerning the regional lymph node in cancer. VII. Thymidine uptake by cells from nodes of breast cancer patients relative to axillary location and histopathologic discriminants

Cells derived from 279 regional lymph nodes (RLNCs) of 62 women with primary breast cancer when evaluated immediately after removal were found to vary significantly relative to their uptake of tritiated thymidine (3HT). The variation was related to the position of nodes in the axilla. Uptake by cells from low axillary nodes was significantly greater than was that by cells from high lying nodes. Consideration is given to the possibility that the increased uptake by low RLNCs reflects the consequences of stimulation by tumor antigen. Relation of the findings to histopathologic discriminants present in nodes and tumors revealed: (a) that the presence of sinus histiocytosis was unrelated to either location of nodes in the axilla or to 3HT uptake by RLNCs, confirming our previous observations indicating no functional or other relationship for that parameter; (b) that while a significantly greater proportion of low than high nodes contained lymph follicles, no correlation existed between 3HT uptake and the presence or absence of follicles in low nodes; and (c) that in general, no relationship existed between nuclear grade or lymphocytic infiltration of tumor and 3HT uptake by RLNCs. The present findings and those reported by others suggest that high axillary nodes in breast cancer patients are more closely related functionally to distant than to low axillary nodes and that despite their anatomical location, they might best be considered biologically and relative to their surgical management in the same category as supraclavicular, cervical, or contralateral axillary nodes. Also, in view of the present findings, the wisdom of removing low axillary nodes in breast cancer operations may be challenged.

Comparison of Concomitant and Sinecomitant Tumor Immunity

Summary Concomitant immunity has been observed in syngeneic rats (Lewis) harboring a progressively growing methylcholanthrene-induced tumor, and in mice (C3H) with spontaneous mammary tumors. Such immunity, as determined by the growth of a challenge of tumor cells from the immunizing tumor, was equivalent to that found in animals whose primary tumors had been removed prior to challenge (sinecomitant immunity). Partial or complete removal of the primary tumor subsequent to challenge failed to alter immunity.

Influence of irradiation of a primary tumor on the labeling index and estrogen receptor index in a distant tumor focus

The present investigation reaffirms our observation that removal of a C3H mouse mammary adenocarcinoma results in a perturbation of tumor cells in a metastatic focus. An increase occurs in the proportion of cells undergoing DNA synthesis (labeling index, LI), and a decrease occurs in the proportion demonstrating estrogen receptor (ER index; ERI). The changes are transient but of sufficient duration and magnitude to produce an increase in the size of a distant tumor. This study was conducted to determine whether cytoreduction of a primary tumor by irradiation would produce a similar change in metastatic tumor cells and whether preoperative radiation would obtund the effect of primary tumor removal. The administration of a maximum tolerated dose of radiation (50 Gy) to a primary tumor produced a significant (p less than 0.001) increase in LI and decrease in ERI of a lesser magnitude than that observed following surgical removal of the primary tumor, but still sufficient to enhance the growth of a metastatic focus. Whereas, there was almost a 50% increase in LI in a metastasis 1 and 3 days following removal of a primary tumor the increase was only 13% three days after radiation. There was a 20% decrease in ERI 3 days following radiation and a 37% decrease at that time following tumor removal. Preoperative irradiation of a primary tumor 1, 3, or 5 days prior to tumor removal, obtunds the increase in LI and decrease in ERI following operation. Radiation the day before surgery was most effective because the changes in a distant focus occurring as a result of the radiation and of the surgery were prevented. The clinical relevance of these observations deserves further consideration.

Studies concerning the regional lymph node in cancer VIII. Effect of two asynchronous tumor foci on lymph node cell cytotoxicity

Results suggest that cytotoxicity by cells from nodes regional to a primary tumor is unique. While a primary tumor was present, A) cytotoxicity was displayed by cells from lymph nodes regional (RLNs) to that tumor, B) cells from nonregional lymph nodes (NRLNs) possessed lesser cytotoxicity which failed to increase in response to a second tumor focus in an area drained by those nodes, and C) the second focus attenuated cytotoxicity of cells from LNs regional to the primary tumor. Following removal of the primary tumor, cells from RLNs rapidly lost cytotoxicity and with passage of time were unable to regain that function in response to a second tumor focus. In contrast, cells from NRLNs demonstrated increased cytotoxicity at any time following removal of the primary tumor when exposed to a second focus. These observations suggest that nodes regional to a distant metastatic focus may be unable to react to it and thus contribute little to the host response generated by the primary tumor. In addition, since nodes regional to a primary tumor manifest diminished cytotoxicity in the presence of a distant tumor focus, tumor cells gaining access and lodging in those nodes subsequent to the development of other metastatic foci are likely to proliferate, resulting in the “positive” lymph node. The findings are in keeping with our contention that host factors responsible for metastases, and perhaps metastases themselves, are at least in part responsible for growth of tumor in RLNs. They also have relevance to the site of administration of specific immunotherapeutic agents and to the significance of the removal of RLNs with a primary tumor.

Further observations on the inhibition of tumor growth by C. parvum with cyclophosphamide. VII. Effect of treatment prior to primary tumor removal on the growth of distant tumor

The present investigations were directed toward determining whether primary tumor manipulation prior to its removal is advantageous for the control of metastases and survival. Studies were carried out to ascertain whether 1) there is justification for delaying surgical removal of a primary tumor to permit preoperative administration of cyclophosphamide (CY) and/or C. parvum (CP) and 2) there is an advantage to administering the immunotherapy directly into a primary tumor. After operation, in all investigations, systemic CP and CY was used. Despite the putative similarity of animals, tumors and treatment regimens there was marked variation in response of tumors to therapy. No benefit was derived from administering preoperative immunotherapy alone. When operation was delayed to employ systemic immuno‐chemotherapy, a slight improvement in the control of distant tumor was noted. The employment of preoperative intratumor immunotherapy led to a greater prolongation of survival and more inhibition of distant tumor growth than did immediate primary tumor removal or the use of preoperative systemic immunotherapy. The results suggest that there may be an advantage to delaying removal of a primary tumor so that it may be employed in therapeutic strategies directed toward control of metastatic disease. Cancer 43:451–458, 1979.